Semi-related, X-37B is heading up next month, and unlike so many missions they're saying a little bit what this one is for. It'll attempt to do some laser communications demonstrations. And... it has a "quantum inertial sensor", as a would-be GPS alternative. https://www.airandspaceforces.com/x-37-eighth-mission-laser-...

The article says it uses magnetic and gravitational field sensing. I don't know how surprised I'd be to hear magnetic field sensing would work: I'd sort of assume there'd be some significant variance day-by-day week-by-week. But getting up there and seeing what you get is probably the way to find out?

If you happen to run three clusters of mag sensors (X,Y,Z on each wing tip and another group of three on a tail boom) then you are measuring a differential field against three points for each axis and sensing the induced ground change from a diurnally fluctuating field .. also getting different results depending on whether you are traveling North to South, East to West, or at some other angle.

Despite all that, maps are still produced of the leveled and normalized ground field (see geomagnetic maps).

Using the geomagnetic as an assist / backup to navigation reduces to something similar to the use of a DTM as an assist - not much use in flat areas, useful in areas with distinct features, easily confused if one valley looks much like another.

The daily flux is an issue, of course, but your software, hypothetically, would be looking at the preserved shape of waves on the ocean despite the fact that the tide is rising and falling all around.